DEBAUCHERY AND ORIGINAL SIN: THE CURRENCY COMPOSITION OF SOVEREIGN DEBT
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
DEBAUCHERY AND ORIGINAL SIN: THE
CURRENCY COMPOSITION OF SOVEREIGN
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
DEBT
Charles Engel JungJae Park
University of Wisconsin-Madison, USA Yonsei University, South Korea
Abstract
We present a model that accounts for the “mystery of original sin” and the surge in local-currency
borrowing by emerging economies in the recent decade. We quantitatively investigate the currency
composition of sovereign debt in the presence of two types of limited enforcement frictions arising
from a government’s monetary and debt policy: strategic currency debasement and default on
sovereign debt. Local-currency debt obligations act as a better consumption hedge against income
shocks than foreign-currency debt because their real value can be affected by monetary policy.
However, this provides a government with more temptation to deviate from disciplined monetary
policy, thus restricting borrowing in local currency more than in foreign currency. Our model predicts
that a country with a less credible monetary policy borrows mainly in foreign currency as a substitute
for monetary credibility. An important extension demonstrates that in the presence of an expectational
Phillips curve, local-currency debt improves the ability of monetary policymakers to commit.
(JEL: E32, E44, F34)
1. Introduction
“Original sin” in the international finance literature refers to a situation in which
emerging economy central governments are not able to borrow abroad in their own
currency. This concept, first introduced by Eichengreen and Hausmann (1999), is
still a prevailing phenomenon for a number of emerging economies, even though the
recent studies by Du and Schreger (2016a,b) and Arslanalp and Tsuda (2014) find that
The editor in charge of this paper was Nicola Pavoni.
Acknowledgments: Charles Engel acknowledges support from the National Science Foundation, grant #
151782 and 1918340. JungJae Park is grateful for the financial support from the Ministry of Education
of Singapore (Academic Research Fund Tier 1 Program NO:FY2017-FRC2-011) and Yonsei university
(New Faculty Research Seed Funding Grant (2021)). This paper has benefited from comments at numerous
conferences and seminars. We thank Manuel Amador, Yan Bai, Javier Bianchi, Galina Hale, Dmitry
Murkhin, and Pablo Ottonello for comments and discussion.
E-mail: cengel@wisc.edu (Engel); jjpark.economist@yonsei.ac.kr (Park)
Journal of the European Economic Association 2022 20(3):1095–1144
https://doi.org/10.1093/jeea/jvac009
c The Author(s) 2022. Published by Oxford University Press on behalf of European Economic Association.
All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
1096 Journal of the European Economic Association
the ability of emerging markets (EMs) to borrow abroad in their own currency has
significantly improved in the last decade.1
A sovereign government faces a temptation to inflate away the real value of local-
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
currency (LC) debt obligations. This temptation to debase or “debauch” the currency
may lead markets to restrict lending in LC debt for some sovereign borrowers. This
temptation has been understood by economists for many years, though the literature
lacks a full model of the dynamic contracting problem in a setting of debasement
and default. Indeed, Keynes (1919) asserted that “Lenin is said to have declared that
the best way to destroy the capitalist system is to debauch the currency.” Keynes
made this point in the context of the debate over debt forgiveness after the World War
I—countries could effectively renege on debt by debauching the currency.2
Building on this idea, a number of studies propose theoretical models that attribute
the predominance of foreign currency external borrowing to EMs’ monetary and fiscal
indiscipline and study its implications for EMs’ economic policy and performance
(e.g. Calvo 1978; Céspedes, Chang, and Velasco 2004; Corsetti and Mackowiak
2005; Jeanne 2005). However, Eichengreen, Hausmann, and Panizza (2005) and
Hausmann and Panizza (2003) find weak empirical support for the idea that the level of
development, institutional quality, or monetary credibility is correlated with the share
of external debt denominated in LC.
Eichengreen, Hausmann, and Panizza (2005) call this empirical finding the
“mystery of original sin”, as EMs seem to suffer from an inherited burden from the past,
regardless of the government’s policies, or their legal or political institutions. They
claim that the original sin problem of EM economies is exogenous to a country’s
economic fundamentals—it is rather related to the structure of the international
financial system.
In this paper, we build a full model of the dynamic contracting problem in a setting
of debasement and default to account for EM economies’ original sin problem and the
“mystery of original sin”. We make the point that contrary to Eichengreen, Haussmann,
and Panizza (2005), the original sin problem can be mainly attributed to a country’s
monetary indiscipline, but that the relationship between endogenous variables, such as
inflation performance, and the currency composition of debt is not straightforward.
For example, borrowers with a low cost of inflation (i.e. countries with less
disciplined monetary policy) prefer a portfolio more weighted toward LC debt because
they can use inflation more easily to make debt repayment state-contingent. However,
the lender may be less likely to offer a portfolio with a large amount of LC debt in such
a scenario because the temptation to excessive currency debasement may be too high
for the borrowers with a low cost of inflation. When there is a lack of commitment
to a sovereign government’s monetary policy, there is a tension between wishes of
sovereign borrowers and lenders regarding how much to borrow and lend in LC, thus
1. For example, Du and Schreger find that the cross-country mean of the share of external government
debt in local currency has increased to around 60% for a sample of 14 developing countries. The countries
in the sample are Brazil, Colombia, Hungary, Indonesia, Israel, South Korea, Malaysia, Mexico, Peru,
Poland, Russia, South Africa, Thailand, and Turkey.
2. See White and Schule (2009) for a discussion of the context of Keynes’s famous statement.
Engel and Park Debauchery and Original Sin 1097
leading to a possibly indirect relationship between inflation performance and the LC
share of sovereign debt.
Using the recent dataset by Arslanalp and Tsuda (2014, 2020), we reconfirm
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
the “mystery of original sin”: The share of external sovereign debt denominated in
domestic currency is not directly related to observable macroeconomic performance.
But we do find evidence that the share of LC debt is related to the decision to use
inflation targeting as a monetary policy, the political stability of the government, and
the depth of capital markets. These variables, in turn, have counterparts in structural
parameters in our model.
We study the currency composition of sovereign debt in the presence of two types
of limited enforcement frictions arising from a government’s monetary and fiscal
discipline: strategic currency debasement and default on sovereign debt. We build
a dynamic general equilibrium model of a small open economy to quantitatively
investigate the implications of these two different enforcement frictions for a
government’s debt portfolio choice.
Our setting is a standard small open economy model with stochastic endowment
shocks, extended to allow a benevolent sovereign government to borrow in both local
and foreign currency. Risk-neutral foreign investors in international financial markets
are willing to lend to the sovereign government any amount, whether in local or
foreign currency, as long as they are guaranteed an expected return at the risk-free
rate prevailing in the international financial markets. Since the real value of repayment
for LC debt can change depending on the ex-post inflation rate (currency depreciation
rate), the foreign investors and the sovereign agree to a LC debt contract that specifies
an inflation rate in each state of the world.
To model inflation targeting—a time-consistent monetary (exchange rate) policy
under commitment—we consider an optimal self-enforcing contract that maximizes
the utility of the representative household in the small open economy and that
prevents the government from breaching the contract in any state of the world.3 Our
approach to modeling the LC debt contract featuring a government’s state-contingent
monetary policy is in line with the optimal dynamic contracting approach to sovereign
borrowing (e.g. Atkeson 1991; Kehoe and Levine 1993; Alvarez and Jermann 2000;
Aguiar, Amador, and Gopinath 2009). This framework is more suitable for studying
a committed time-consistent monetary policy than the Markov equilibrium approach,
which quantitative incomplete market sovereign default models adopt (e.g. Aguiar and
Gopinath 2006; Arellano 2008; Ottonello and Perez 2019). The latter is better suited
for studying discretionary monetary policy.
Why do we connect the ability to borrow in LC with the adoption of an
inflation-targeting monetary policy? Inflation targeting is a form of monetary policy
commitment, whose analog in our framework is the willingness of the sovereign
borrower to adopt a LC contract with state-contingent inflation rates. Svensson (1997)
points out that a strong commitment to a systemic monetary policy through inflation
3. We use the term “inflation targeting” throughout, though in fact, the model incorporates monetary
policy that is akin to price-level targeting. That is, high inflation in some states of the world is balanced by
low inflation in other states in the optimal contract.
1098 Journal of the European Economic Association
targeting can mimic the optimal inflation contract offered to an independent central
banker suggested by Walsh (1995). This optimal contract can also be thought of as a
credible time-consistent optimal monetary policy in a similar vein to Lucas and Stokey
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
(1983) and Chang (1998). Moreover, the structure of the LC debt contract in the model
mimics the inflation targeting framework in the real world: It features a target inflation
rate with flexible inflation rate bands.
Our model predicts that the optimal contract for LC debt allows the government
to inflate away a certain fraction of LC debt in times of bad income shocks but
asks for currency appreciation in times of good shocks as a compensation for the bad
times. However, due to the limited enforcement constraint arising from a government’s
temptation to inflate away LC debt, the borrowing limit for LC is endogenously
constrained, thus restricting the consumption hedging benefit of LC debt. Meanwhile,
the enforcement constraint arising from the option to fully default on its debt mainly
determines the endogenous borrowing limit for foreign-currency debt. These two
enforcement frictions combine to generate an endogenous debt frontier, determining
the maximum amount of debt in each currency.
In our model, as well as in reality, a sovereign’s default on debt is generally more
costly for its economy than excessive currency debasement. Costly default provides
foreign-currency debt with more credibility than LC debt, thus allowing the sovereign
to borrow more in foreign currency. This extra credibility from costly default makes
foreign-currency debt valuable, whereas the state-contingency inherent in LC debt
makes LC debt valuable: A sovereign with less disciplined monetary policy borrows
mainly in foreign currency as a substitute for monetary credibility. Thus, we see a mix
of foreign- and local-currency debt in equilibrium.
We calibrate the model to a panel of nine EM economies, which adopted inflation
targeting in the last decade. The quantitative results show that a country with more
disciplined monetary policy—represented by a country with a high cost of inflation
in our model—can borrow more in both foreign and local currency. More disciplined
countries can borrow more in LC, which provides a better consumption hedge. A
country with less disciplined monetary policy wants to borrow more in LC but is
restricted to borrow mainly in foreign currency due to the enforcement constraint. We
also conduct a dynamic analysis to account for the recent increase in EM economies’
LC borrowing. We find that the increase in a country’s monetary credibility, represented
by the cost of inflation in our model, almost exclusively accounts for the recent increase
in EM’s LC borrowing.
The currency composition of debt and variables such as the volatility of inflation
are all endogenous. They depend on the economy’s characteristics, such as the degree
of patience and risk aversion, and the cost of default and inflation, as well as its income
shock process. The cross-country analysis, in which we calibrate the model for each
of nine countries, shows that there is no simple monotonic relationship among these
variables, which may account for the lack of a clear-cut link between the currency
composition of the external portfolio and endogenous macroeconomic variables.
We also consider a version of the economy in which policymakers face an
expectational Phillips curve, which allows the possibility of using monetary policy
to smooth output fluctuations. However, monetary authorities are not endowed with
Engel and Park Debauchery and Original Sin 1099
the power to commit to a policy plan. If the economy can only borrow in foreign-
currency denominated debt, or if it is in financial autarky, then monetary policy is
discretionary. But when a country can obtain a contract to borrow in LC, the value
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
of that contract is a commitment device that allows the policymaker to stick to a
state-contingent pre-announced monetary policy.
In Section 2, we revisit the mystery of original sin problem by Eichengreen,
Hausmann, and Panizza (2005) and Hausmann and Panizza (2003) to empirically
investigate the determinants of the original sin problem. In Section 3, we present our
formal model. In Section 4, we calibrate the model to a panel of nine EM economies and
conduct both dynamic and cross-country analysis to investigate what drives observed
variation in LC borrowing in the cross-section and across time. Then, Section 5 presents
the model with a Phillips curve.
Related Literature. Our work builds on the intuition from the classical argument that
attributes the predominance of foreign-currency debt in international financial markets
to a lack of monetary credibility. A government’s strategic debasement of the real value
of debt can pose a significant obstacle to issuing LC debt (Calvo, 1978; Kydland and
Prescott, 1977).
Bohn (1990) builds a model in which governments can only commit to repayment
of nominal sums and have an incentive to inflate away debt. In Bohn’s setup, some
domestic-currency debt is sustainable because the government bears an exogenous
cost of inflation. In more recent work, Ottonello and Perez (2019) study the currency
composition of sovereign debt in a dynamic general equilibrium model of a small open
economy with a government with limited commitment to monetary and debt policy.
However, Ottonello and Perez (2019) study discretionary policy rather than committed
time-consistent policy with the same frictions. In both models, the original-sin regime
in which governments can borrow only in foreign currency arises only as the special
case in which the cost of inflation is zero. In practice, there must be a fairly high cost
of inflation internally to underpin realistic levels of domestic currency borrowing in
these models.4
Phan (2017) examines an Eaton and Gersovitz (1981) style model with local and
foreign-currency borrowing subject to strategic default and debasement risk. That
paper posits a trigger strategy for the borrower that will support borrowing in LC and
shows that equilibrium LC debt can be sustained even if the punishment for default
or complete debasement of LC debt allows for the country to save in foreign-currency
assets. It thus offers a possible resolution to the Bulow and Rogoff (1989) puzzle
concerning the sustainability of sovereign debt when exclusion is the only punishment
for default, but, in common with Bohn and Ottonello–Perez, it cannot account for
original sin.
In our model, lenders recognize that the sovereign borrower has an incentive to
inflate away the debt and that this option to inflate is more valuable to the borrower
4. See also Du, Pflueger, and Schreger (2020), who consider a two-period framework with risk averse
lenders and compare the cases of a borrowing sovereign without and with commitment.
1100 Journal of the European Economic Association
when, for example, it is suffering from low output or has high debt obligations.
The lender and the sovereign negotiate a contract that allows for more inflation in
circumstances such as this. In that sense, inflation is akin to an “excusable default”, as
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
in Grossman and van Huyck (1988).
Our work draws on, and is closely related to, models with optimal dynamic
contracts in the presence of commitment problems. Atkeson (1991), Kehoe and Levine
(1993), Zhang (1997), Alvarez and Jermann (2000), and Bai and Zhang (2010) are
the closest analogs. These studies show that borrowing limits arising from the limited
enforcement problems can cause significant distortions to allocations of an economy.
As in these studies, there are no equilibrium breaches of the contract in our model.
Our starting point resembles Eaton and Gersovitz (1981), Zhang (1997), Aguiar
and Gopinath (2006), Arellano (2008), and Bai and Zhang (2010) in that we assume that
only bonds that are nominally non-state-contingent can be traded. As in those papers,
we do not derive this limitation endogenously, and instead appeal to the real-world
observation that sovereign debt is typically not explicitly state contingent. However,
our paper is unique in that it recognizes the two ways in which the debt repayments
may be state contingent—because of debasement and outright default. Thus, our model
shares some of the features of both strands of literature—optimal contracts but with
debt that has some, but not full, state contingency.
2. Empirical Analysis
2.1. Mystery of Original Sin Revisited
Eichengreen, Hausmann, and Panizza (2005) and Hausmann and Panizza (2003) find
that empirically, there seems to be very little link between the share of external debt
denominated in LC and variables such as the volatility of inflation or measures of
economic development. These studies find that only the absolute size of the economy,
proxied by its GDP, is robustly correlated with original sin. They call their finding the
“mystery of original sin” and claim that the original sin problem of EM economies is
exogenous to a country’s economic fundamentals.
Ogrokhina and Rodriguez (2018) reexamine the economic determinants of LC
debt. They confirm the findings that typical measures of inflation performance have
limited power in accounting for which countries are able to issue LC government
debt. However, they introduce some new variables that do partially unlock the mystery
and which are consonant with the model we present. In particular, they find evidence
that countries that adopt inflation targeting have been more successful at issuing
sovereign debt denominated in LC—and that effect is independent of actual inflation
performance.
These studies have relied on a less-than-perfect measure of foreign-held LC
sovereign debt, relying primarily on measures of LC debt issued in foreign markets.
Two new studies have tried to carefully measure all sovereign debt held outside a
country’s borders, including by the currency denomination of the debt, irrespective of
whether the debt was issued abroad or within the sovereign. The studies by Du andEngel and Park Debauchery and Original Sin 1101
Schreger (2016b) and Arslanalp and Tsuda (2014, 2020) have a narrower range of
countries coverage than the previous studies, but the data they produce more neatly
lines up with the variable of interest in our study, which is the amount of sovereign
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
debt held abroad that is denominated in LC. The measures in the two studies are
highly correlated. We use Arslanalp and Tsuda’s data, which is publicly available and
periodically updated.
The country coverage of the dataset is spotty prior to 2010, so we cannot undertake
an extended time series analysis. Eliminating from the study countries such as China
and India, for which only a small fraction of their sovereign debt is held by foreigners,
whether in LC or not, leaves 19 countries to investigate. Figure 1 displays the evolution
of the share of their debt held externally that is denominated in LC.
Many of the countries in the sample were undergoing a shift in the early years
of the 2000s from a regime in which little or none of their sovereign debt was in
LC, toward a regime in which a significantly larger fraction was. By the end of the
sample, for example, around 90% of Malaysia’s and Thailand’s debt was in LC. Other
countries, such as Brazil, Mexico, and South Africa, saw significant increases. On the
other hand, some countries saw very little increase in the share of debt in LC.
One likely explanation for the increase is that most of the countries in the sample
officially adopted inflation targeting as their monetary policy stance in the late 1990s or
early 2000s. There were twelve countries in the sample in this category: Brazil (adopted
in 1999), Colombia (1999), Hungary (2004), Indonesia (2005), Mexico (2001), Peru
(2002), Philippines (2002), Poland (1998), Russia (2008), South Africa (2000), Turkey
(2006), and Thailand (2000). In addition, Malaysia was understood to be an inflation
targeter, but they did not adopt that position officially. Only six of the countries did
not adopt inflation targeting: Argentina, Bulgaria, Latvia, Lithuania, Romania, and
Ukraine. In all these cases, the share of external debt denominated in LC remained
very low throughout the time period.
However, a complication to our empirical analysis is that several countries in the
sample were committed to joining the euro area. In fact, Latvia (in January 2014) and
Lithuania (in January 2015) did adopt the euro. The data includes euro-denominated
debt as foreign-currency debt, but that is an ambiguous proposition for these countries.
Romania, Bulgaria, Hungary, and Poland are also in this group, though none has yet
adopted the euro. However, as noted, Hungary and Poland continued to devote their
monetary policy toward inflation targeting. In our regression analysis, then, we will
classify those two countries as pursuing inflation targeting, and we will group the
countries that do not target inflation separately into those looking to join the euro area
and the two remaining countries, Argentina and Ukraine. Although the fraction of LC
sovereign debt held abroad is low for all the countries that do not target inflation, the
reason might be different for the European countries versus the other two.
Figure 2 presents scatterplots of the share of externally held debt in LC against
measures of inflation performance. The share of home currency debt is measured as
the average over the 2010–2019 period. The first panel of the figure illustrates the
relationship between average inflation rates (for the 2000–2019 period) and LC shares,
while the second panel graphs the standard deviation of annual inflation (again for1102 Journal of the European Economic Association
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
F IGURE 1. Share of debt in LC.
2000–2019) against LC shares. Figure 2 illustrates the mystery of original sin—there
appears to be very little relationship between inflation performance and the share
of debt held abroad that is denominated in LC. We look at the share of debt over the
2010–2019 period because the currency composition of debt takes some time to adjust,
even after a country has adopted inflation targeting. There might be a lag between the
implementation of the policy and when markets conclude that the policy commitmentEngel and Park Debauchery and Original Sin 1103
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
F IGURE 2. LC share and inflation volatility. LC share refers to the share of externally held sovereign
debt denominated in LC. For LC share, we use the period average from 2010 to 2019. For the inflation
related variables, we use the sample period from 2000 to 2019.
is credible, but also some foreign-currency debt is longer term, so it takes time for its
share of total externally held debt to decline.
Table 1 presents panel estimates of the determinants of the share of externally held
foreign debt denominated in LC for the 19 countries, using annual data 2010–2019.
Inevitably, because of the paucity of data, it is not judicious to include a large number
of explanatory variables in a single regression. The reported regressions consider nine
independent variables in various combinations: A dummy variable for the countries that
did not adopt inflation targeting, a dummy variable for the countries that were aiming
to adopt the euro, annual inflation, the Chinn-Ito measure of capital account openness,
the World Bank’s World Development Indicator measure of financial development,
domestic credit relative to the non-government component of GDP, log of total GDP,
log of GDP per capita, and the indexes of government effectiveness and political
stability from the World Bank’s World Governance Indicator.
The countries that did not adopt inflation targeting had significantly lower shares
of LC debt. The apparent upswing in LC sovereign debt appears to be related to the
adoption of inflation-targeting monetary policy, as Ogrokhina and Rodriguez (2018)
found in their dataset of foreign-issued debt. Additionally, Bulgaria and Romania
issued little domestic currency debt and instead sold debt to foreigners denominated
in euros, consistent with their aim to join the euro area.
Inflation performance itself has little explanatory power, which is consistent with
the mystery of original sin documented by Eichengreen et al. (2005) and Hausmann and1104
TABLE 1. Determinants of LC foreign-held sovereign debt. Dependent variable: share of externally held sovereign debt denominated in LC.
Panel ordinary least squares,
annual, 2010–2019 (1) (2) (3) (4) (5) (6) (7) (8)
Inflation non-targeter 0.1103 0.4048 0.4202 0.3630 0.3745 0.4183 0.3070 0.4271
(0.1137) (0.0802) (0.0769) (0.0757) (0.0877) (0.0769) (0.0876) (0.0744)
Joining euro area 0.0968 0.4291 0.4349 0.3786 0.2381 0.4491 0.4829 0.4887
(0.1142) (0.0740) (0.0810) (0.0674) (0.1349) (0.0753) (0.1020) (0.0786)
Inflation 0.0006 0.0022
(0.0014) (0.0021)
Capital account openness 0.0288 0.0315
(0.0753) (0.0659)
Domestic credit/GDP 0.3062 0.2473
(0.0852) (0.1161)
ln(GDP) 0.1564 0.0874
(0.0532) (0.0534)
ln(GDP) per capita 0.1255 0.0660
(0.1140) (0.0677)
Government effectiveness 0.2612 0.2221
(0.1072) (0.1017)
Political stability 0.0333 0.0732
(0.0319) (0.0457)
N 166 184 166 184 184 184 184 184
Journal of the European Economic Association
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022TABLE 1. Continued
(9) (10) (11) (12) (13) (14) (15) (16)
Inflation non-targeter 0.2995 0.4198 0.1951 0.3957 0.2406 0.3921 0.2877 0.4203
(0.1041) (0.0787) (0.0929) (0.0701) (0.1038) (0.0895) (0.0923) (0.0763)
Joining euro area 0.4788 0.4875 0.4404 0.4348 0.3036 0.3721 0.4881 0.4906
(0.1012) (0.0839) (0.1080) (0.0638) (0.1515) (0.1697) (0.0945) (0.0752)
Inflation 0.0011 0.0003 0.0011 0.0007 0.0006 0.0000 0.0014 0.0003
(0.0008) (0.0015) (0.0014) (0.0019) (0.0013) (0.0020) (0.0014) (0.0020)
Capital account openness 0.0146 0.0279
(0.0695) (0.0658)
Domestic credit/GDP 0.2780 0.2042
(0.1046) (0.1032)
ln (GDP) 0.0845 0.0511
Engel and Park Debauchery and Original Sin
(0.0480) (0.0411)
ln (GDP) per capita 0.0089 0.0107
(0.0783) (0.0846)
Government effectiveness 0.1897 0.2459 0.2199 0.2171
(0.1042) (0.0918) (0.1015) (0.1133)
Political stability 0.0629 0.0530 0.0643 0.0700
(0.0396) (0.0402) (0.0474) (0.0517)
N 166 166 184 184 184 184 184 184
Notes: The table reports the OLS estimates of the coefficient of the panel regression listed. The data are annual, 2010–2019. There is an intercept term in each regression, not
reported. The 19 countries in the sample include the inflation-targeting countries of Brazil, Columbia, Hungary, Indonesia, Malaysia, Mexico, Peru, Philippines, Poland, Russia,
South Africa, Turkey, and Thailand. The countries dummied as aiming to join the euro area are Bulgaria and Romania. The two dummied as non-inflation targeters are Argentina
and Ukraine. Inflation is CPI inflation. Capital account openness is the Chinn-Ito measure. Domestic credit to the private component of GDP is from the World Development
Indicators of the World Bank. The measures of government effectiveness and political stability are from the World Bank’s World Governance Indicators. OLS standard errors in
parentheses. , , and indicate that the alternative model significantly different from 0 at 10%, 5%, and 1% significance level, respectively, based on standard normal critical
values for the two-sided test.
1105
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 20221106 Journal of the European Economic Association
Panizza (2003). We have also included measures for the size of the country (log GDP)
and the level of income per person (log of GDP per capita). The former measure had
explanatory power in the earlier studies, but we generally find that it does not, except in
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
the first specification of Table 1. As in the earlier studies, income per person as a proxy
for economic development is not helpful in explaining the currency denomination of
sovereign debt.
We do find a few measures that have success in explaining the ability to sell
sovereign debt abroad that is denominated in LC. We find that countries with greater
financial development and countries with greater government effectiveness are better
able to issue domestic-currency debt.
Overall, our findings are similar to the earlier studies of original sin. We have, in
Table 1, reported results for some of the variables that were successful in accounting for
LC debt, but we looked at many more. In panel regressions, we considered measures
such as the total government debt relative to GDP, trade openness, and other measures
of governance. We also did cross-sectional regressions and considered more measures
of inflation performance, such as average inflation, the standard deviation of inflation,
and the maximum inflation rate over various sub-periods. We also considered as
explanatory variables the classification of the exchange-rate regime from Ilzetzki,
Reinhart, and Rogoff (2019) and the Fernandez et al. (2016) measure of capital control
intensity. None were useful explanatory variables.
We note that variables that do appear to predict that a country can escape original
sin and issue significant amounts of LC debt—commitment to inflation targeting,
financial development, and government effectiveness—have analogs in the form of
deep parameters in our theoretical model. In Section 4, we calibrate the model
and examine whether the calibrated parameters are consistent with their empirical
counterparts.
3. The Model Economy
We consider a standard small open economy model, extended to allow a sovereign
government to borrow in both local and foreign currency from foreign lenders in
international financial markets. Time is discrete (t D 0, 1, 2, 3. . . ) and runs forever.
Before the income shock is realized at period 0, the sovereign attempts to arrange a LC
debt contract with the foreign lenders. If the sovereign and foreign lenders successfully
agree on the terms of a LC debt contract, then the small open economy can borrow in
both currencies thereafter. On the other hand, if the sovereign and foreign lenders fail
to agree, then the economy may borrow only in foreign currency thereafter.
The representative household receives stochastic endowment shocks every period
and has preferences given by
1
X
E0 ˇ t Œu.ct / C.t /;
N (1)
tD0Engel and Park Debauchery and Original Sin 1107
where ˇ denotes the time discount factor, ct consumption, t the gross inflation rate
at period t (i.e. Pt /Pt1 ), and N the target inflation rate of the country. The period
utility function u(.) is differentiable, strictly increasing, strictly concave, and satisfies
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
the standard Inada conditions. Following Barro and Gordon (1983), we introduce a
cost of inflation in the form of utility loss C.t /, N which is differentiable and is
symmetric around the target inflation rate ; N any deviation in inflation rates from the
target inflation rate incurs utility loss. The sovereign government is benevolent and
makes borrowing, default, and debasement decisions to maximize the welfare of this
economy.
There is one tradable consumption good in this economy. The random income
shock yt has a finite support Y D fy1 ,y2 , . . . , yN g and follows a Markov process with a
transition function Pr.ytC1 jyt /. The history of the income shock is denoted by st . Let
Pt and Pt be the prices of the consumption goods in the home (i.e. the small open
economy) and foreign countries, respectively. The budget constraint in nominal terms
is given by
Pt ct C St Pt btC1 C Pt btC1 D Pt yt C R St Pt1
for for
loc
bt C it Pt1 btloc ; (2)
for
where St is the exchange rate, bt 0 foreign-currency debt, btloc 0 LC debt, it the
gross interest rate on LC debt, and R the constant gross risk-free rate prevailing in
for
international financial markets.5 We assume that b0 and b0loc are initially given, P1
D 1 and y1 is given with Pr.y1 / D 1. We also assume that the law of one price holds
and the foreign price Pt is normalized to be one, so that Pt D Pt St D St . Then the
budget constraint for the economy, conditional on the sovereign government rolling
over its debt by following the terms of contract, is given in real terms by
it btloc
D yt C R bt C
for for
ct C btC1 C btC1
loc
: (3)
t
When the government does not breach the contract, it solves a portfolio problem
between local- and foreign-currency debt to maximize the social welfare of the
economy. Finally, we impose the natural debt limit following Aiyagari (1994) given
by
for
btC1 C btC1
loc
D; (4)
where D D y=.R 1/ and y are the lowest income shock.
5. Since we investigate the currency composition of two types of sovereign debts, we don’t allow the
government to accumulate assets. bloc > 0 is ruled out because it would not be plausible to assume that the
foreign lenders issue debt in the currency of the small home country. The small open economy could not
punish a large lender such as the United States either for default or debasement. We don’t allow bfor > 0
because we want to rule out an equilibrium in which the economy borrows only in local currency and save
only in foreign currency at the same time. In any case, the no accumulation constraint is not binding in the
simulations. This is because when ˇ (1 C r ) < 1, the economy wants to borrow to achieve a front-loaded
consumption profile, but due to debasement risk, the economy cannot borrow in local currency as much as
it wants. In this case, the economy needs to rely on foreign currency debt to satisfy its borrowing need.1108 Journal of the European Economic Association
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
F IGURE 3. Two types of breaches of contract.
The government can breach the debt contract in the following two ways: First,
the government can fully default on its debt denominated in both local and foreign
currency simultaneously. Selective default on a certain type of debt is not allowed in
our model, consistent with practices in sovereign debt markets and the theory in the
sovereign debt literature.6 Second, the government can debase its currency more than
required in the LC contract, the terms of which will be specified in detail later. Thus,
our model features two types of enforcement (commitment) frictions arising from a
government’s monetary and fiscal indiscipline: strategic default and debasement.
When the government fully defaults on its debts, the economy enters financial
autarky, during which it loses access to international financial markets. When the
government breaches the contract through excessive currency debasement, the country
is restricted to borrow only in foreign currency as a punishment, thus entering the
original sin regime. When the government in this regime defaults on its foreign-
currency debt, the economy also enters financial autarky. Figure 3 summarizes the two
different types of breaches of the debt contract and their consequences.
In the benchmark model, we assume temporary financial autarky as a punishment
for outright default and permanent original sin regime as a punishment for excessive
debasement. However, even if we assume a temporary original sin regime as a
6. See Broner, Martin, and Ventura (2010) for a theoretical study on this problem.Engel and Park Debauchery and Original Sin 1109
punishment for the excessive debasement, our main results carry through (see Online
Appendix D).
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
3.1. LC Debt Contract
Foreign lenders in competitive international financial markets are risk-neutral and have
deep pockets. There are two types of lenders: lenders who lend in LC and those in
foreign currency. Both are willing to lend to the sovereign government any amount,
whether in local or foreign currency, as long as they are guaranteed an expected return
of the gross risk-free rate R . Even if LC debt is non-contingent in nominal terms with
a gross nominal interest rate it , depending on the government’s ex-post choice of the
inflation rate t (or equivalently currency depreciation rate), the real rate of interest
on LC debt it / t can differ. We consider the following recursive contract for LC debt,
which consists of two components: a nominal gross interest rate it and state-contingent
inflation rates in the next period t .7
for
it D I bt ; btloc ; yt1 ; (5)
for
t D … bt ; btloc ; yt1 ; yt : (6)
for
When the sovereign government borrows bt and btloc in foreign and local currency
in period t 1, the contract specifies a nominal gross interest rate it on LC debt btloc .
Moreover, the contract asks for an inflation (currency depreciation) rate that depends
on the realization of yt in period t.
Since the foreign investors who lend in LC must be guaranteed an expected return
of a gross risk-free rate R for the LC debt, we have the following zero-profit condition
on the contract:
for loc
X i
t t b ; b t ; yt1
R D P r yt jyt1 : (7)
for
yt t bt ; btloc ; yt1 ; yt
Note that …() contains yt1 as well as yt because of the persistent income shock
process Pr.yt jyt1 / in equation (7). In the next section, we specify an optimal self-
enforcing LC contract subject to the lenders’ zero-profit condition to model inflation
targeting as a state-contingent monetary policy under commitment. Due to the zero-
profit condition (equation (7)), a currency depreciation (or equivalently t above the
target inflation rate )
N at a certain state in the contract must be accompanied with a
currency appreciation (i.e. t below )
N at other states.
7. It is restrictive to consider only recursive contracts even though this particular form of the recursive
contract imposes no additional restrictions. When the contract is not Markov (i.e. the contract is history
dependent), it could take many different forms. For example, investors could propose a contract that asks
for additional interest premiums on the bonds issued by sovereigns that have a history of breaching the
contract.1110 Journal of the European Economic Association
This LC contract featuring the government’s state-contingent monetary policy
mimics the inflation targeting framework in the real world: Inflation targeting usually
takes a form of a target inflation rate N with flexible inflation rate bands. In this context,
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
the breach of the LC contract through excessive debasement can be interpreted as a
sovereign’s breach of the self-announced inflation rate bands by conducting a reckless
monetary policy that leads to high or hyperinflation in the economy8 ; once the sovereign
decides to deviate from the contracted inflation rate, it will choose a very high inflation
rate to maximize the reduction in the real value of the LC debt.
On the other hand, the foreign lenders charge the gross risk-free rate R on the
foreign-currency debt as typical of a standard small open economy model featuring
non-contingent debt. From now on, xt denotes the vector of state variables at period t,
for
which consists of .bt ; btloc ; yt1 ; yt /.
3.2. Discussion of Main Assumptions
Two Types of Punishment. Our paper derives a reputational equilibrium as in Eaton
and Gersovitz (1981), but a sovereign government in our model has two types of
reputation: repayment and monetary reputation. If a sovereign defaults on its debt,
whether it is in foreign or local currency, then it loses repayment reputation, banning it
from international financial markets. As with the sovereign debt literature and empirical
evidence, we assume that a defaulting country enters financial autarky temporarily,
regaining access to international financial markets after a stochastic number of periods.
On the other hand, if the sovereign breaches the contract through excessive currency
debasement, then its monetary reputation is lost, so lenders would not lend any amount
in LC (i.e. the country enters the original sin regime). Note that lenders would lend in
foreign currency to the sovereign that lost its monetary reputation but not the repayment
reputation. This implies that the LC lenders cannot impose financial autarky on a
government that loses only its monetary reputation after excessive debasement: The
foreign-currency lenders will not cooperate with the LC lenders in a punishment of
financial autarky.9
Reinhart and Rogoff (2009) document the prevalence of high inflation periods
for many countries throughout history, and that in the aftermath of high inflation
or hyperinflation, these countries generally experienced a huge shift toward the use
8. Unlike default, which is a discrete event, excessive debasement—deviation from the inflation target—
is continuous, so upon even an infinitesimal deviation from the inflation target, the punishment must kick
in. However, in practice, when the sovereign decides to breach the contract through excessive debasement,
it would choose a finite level of inflation substantially above the inflation target to equate the benefit of
inflating away the local currency debt and the cost of inflation at the margin. If inflation costs are zero,
then the country will generate an infinite inflation. Upon observing this high inflation, lenders would stop
lending in local currency to the sovereign, which has lost monetary reputation.
9. It is common for courts in unsecured lending countries to hinder the defaulting a sovereign’s new
issuance of debt by impairing the rights of new lenders until the sovereign reaches a settlement with the
original lenders. But in the case of inflation, the nominal debt is repaid, so there are generally no legal
grounds for prohibiting further lending to a high inflation country.Engel and Park Debauchery and Original Sin 1111
of foreign currency for transactions and borrowing (i.e. “dollarization”) because the
governments’ monetary credibility had been lost. Original sin is a specific case of
dollarization (Yilmaz 2006). The punishment of original sin for excessive debasement
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
is motivated by this historical fact, as the punishment of financial autarky for outright
default is motivated by the historical facts regarding sovereign defaults (Eaton and
Gersovitz 1981).
Unlike temporary financial autarky as a punishment for outright default, we
assume a permanent original sin regime for the country that has breached the contract
through excessive debasement because there is no historical episode in which any
county suffered from a temporary original sin regime. Since the international financial
liberalization in the late 1970s, only a handful of countries have been able to borrow
externally in LC (the United States, the United Kingdom, Japan, Switzerland, and
Germany), and the vast majority of countries were not able to borrow in their domestic
currencies until recently.10 EM economies’ recent escape from the original sin should
be thus viewed not as their regaining access to the ability to borrow in LC but as
their having overcome the original sin problem after adopting the inflation targeting.
Proposition 2 in Section 2 shows that only countries with certain characteristics are
able to borrow in LC.
The punishment for excessive debasement—forcing the sovereign into the original
sin regime—is generally less costly to the economy than the punishment of financial
autarky for outright default. Corollary 2 in Section 2 shows that if both outright default
and excessive debasement are equally punished with financial autarky, then LC debt
has no debasement risk, so the sovereign can always borrow in LC. That is, we would
not see any economy suffering from original sin in equilibrium with equal punishment,
an implication that is not consistent with the fact that most EM economies still borrow
only in foreign currency. The corollary provides additional support for our choice of
different punishments for outright default and excessive debasement.
LC Debt Contract with State-Contingent Inflation Rates. Unlike Atkeson (1991),
Kehoe and Levine (1993), and Alvarez and Jermann (2000), which assume a full set
of state-contingent claims traded internationally, we make a more realistic assumption
that the asset markets are incomplete: The set of assets that the government can issue
is restricted to non-contingent bonds in nominal terms, whether denominated in local
or foreign currency.
Grossman and van Huyck (1988, 1993) point out that throughout history,
international loans, while non-contingent in nominal terms, are state contingent through
frequent rescheduling of repayments, often via currency debasement. Grossman (1990)
and Grossman and van Huyck (1993) suggest that the UK and the US’s currency
appreciation on returning to gold at the prewar parity after the Napoleonic War
10. Bordo, Meissner, and Redish (2005) show historical cases in which how several countries including
the United States have overcome the original sin problem in the early 20th century. Except for the United
States and the United Kingdom, virtually all countries suffered from the “original sin” regime throughout
the history before the 1970s, and even the US external sovereign debt had implicit or explicit gold clauses
until 1933.1112 Journal of the European Economic Association
(United Kingdom), the Civil War, and the World War II (United States) were motivated
to establish a trustworthy reputation with foreign lenders to maintain access to future
loans for the next war.
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
Even under the incomplete asset market structure, however, a sovereign government
can work out a time-consistent optimal state-contingent monetary and debt policy in a
similar spirit to Lucas and Stokey (1983) and Chang (1998). The government can then
announce the monetary and fiscal policies to foreign lenders and ask them to agree to
the terms of contract featuring the policy. As long as foreign lenders get an expected
return of R , risk-neutral lenders would accept the contract.
We interpret the sovereign borrower’s adoption of a LC contract with state-
contingent inflation rates as a commitment to inflation targeting. Svensson (1997)
points out that a stronger commitment to a systemic monetary policy through inflation
targeting, with increased transparency through communication with the public and
increased accountability of the central bank, can mimic the optimal inflation contract
offered to an independent central banker suggested by Walsh (1995).
In our model, targeting the exchange rate and targeting inflation (or, more precisely,
the price level) are equivalent, given the assumption of the law of one price and the
exogenous nominal price in the lending country. In the real world, those policies have
often been perceived as being different. Exchange-rate targeting has often involved
capital controls and sterilized intervention, and monetary policy has not been rigorously
devoted to stabilizing inflation. The real appreciation that many exchange-rate peggers
experienced was mostly not a result of rapid productivity gains in the traded sector that
led to increases in the relative price of nontraded goods but was more attributable to
controls that fixed the exchange rate while allowing CPI inflation to continue. This was
often reflected in black-market exchange rates, which diverged from official pegged
rates.11 We interpret the adoption of inflation targeting as the recognition that the old
policies, which did not target inflation, were untenable politically and in terms of
economic welfare.
3.3. Value of Debasement
Due to the limited commitment (enforcement) of monetary policy, the sovereign
government can debase its currency excessively at any time by choosing a higher
inflation rate than (xt ) called for in the contract, in order to inflate away a certain
fraction of LC debt. When the government breaches the contract through excessive
debasement, the country is restricted to borrowing only in foreign currency thereafter
as a punishment. That is, the country enters the regime of original sin or foreign-
currency borrowing. The sovereign in the original sin regime cannot commit to any
monetary and debt policy. In the original sin regime, we consider a Markov perfect
equilibrium, in which the sovereign conducts discretionary monetary and debt policy
11. See, for example, Edwards (1989).Engel and Park Debauchery and Original Sin 1113
and defaults on its debt in equilibrium, as in the standard quantitative sovereign debt
literature (Aguiar and Gopinath 2006 and Arellano 2008).12
The value of debasement is given by
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
for
V debase bt ; btloc ; yt1 ; yt I it ; .xt /
; (8)
N C ˇEt V o btC1 ; ytC1
for
D max Œu.ct / C.t /
for
t ¤.xt /;ct ;btC1
subject to the budget constraint:
i b loc
ct C q btC1 ; yt btC1 D yt C R bt C t t ;
for for for
(9)
t
for
btC1 D: (10)
Vo () denotes the value of the option to default in the original sin regime after the
debasement. At the time of excessive debasement, the sovereign still needs to pay back
fully the nominal amount of its LC debt, but the debasement reduces the real repayment.
for
The sovereign rolls over its debts—both local- and foreign-currency debts—with btC1
for
at the discount bond price q.btC1 ; yt /, which will be defined in the next subsection.
3.4. Value of Foreign-Currency Borrowing (Original Sin Regime)
If a sovereign and lenders fail to agree on terms of LC debt at period 0 (i.e. there exists
no sustainable LC debt contract (see Definition 1)), or if the sovereign breaches the
LC contract through excessive debasement at a later period, then the economy must
borrow only in foreign currency thereafter.
In the original sin regime, the sovereign can default on its debt at any time after
comparing the values of default and no default. Upon default, the economy enters
financial autarky, in which it temporarily loses access to foreign-currency borrowing
from international financial market.
The value of the option to default when the sovereign has access only to foreign-
currency borrowing is given by
n o
V o bt ; yt D max V for bt ; yt ; Vfor .yt / ;
for for def
(11)
def
where Vfor denotes the value of default, when the economy enters financial autarky
from the foreign-currency borrowing regime. The value of foreign-currency borrowing
for the case of no default is given by
N C ˇEt V 0 btC1 ; ytC1 ;
for for
V for bt ; yt D max Œu.ct / C.t / (12)
for
ct ;t ;btC1
12. In the previous version of this paper, we considered an equilibrium in which the sovereign does not
default in equilibrium in the original sin regime. Allowing for equilibrium default does not result in any
qualitative changes in simulation results of our model.1114 Journal of the European Economic Association
subject to the following constraints:
for for for
ct C q btC1 ; yt btC1 D yt C bt ; (13)
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
for
btC1 D: (14)
for
The bond pricing function q.btC1 ; yt / is defined as follows:
1 X
for for def
q btC1 ; yt D Pr ytC1 jyt I V for btC1 ; ytC1 > Vfor ytC1 ;
R
ytC1 2Y
(15)
where I is an indicator function that takes on the value of one if the condition inside the
parenthesis holds, and zero otherwise. With no LC debt on hand, the optimal t D . N
3.5. Value of Default
Upon default, the economy enters financial autarky for a stochastic number of periods,
during which the economy loses access to international financial markets, and the
economy suffers a drop in income. There are two types of financial autarky: one
resulting from default in the original sin regime (the economy has lost both monetary
and repayment reputation) and the other from default when the sovereign has a LC
contract (the economy has lost only the repayment reputation). In both cases, the
countries re-enter their respective credit markets with an exogenous probability , and
start with zero debt.
The value of default when the economy enters financial autarky from the original
sin regime is given by
N C ˇ.1 /Et Vfor .ytC1 / C ˇEt V o .0; ytC1 /;
def def
Vfor .yt / D maxŒu.ct / C.t /
t
(16)
ct D h.yt /; (17)
where h(yt ) < yt . h(yt ) represents a decrease in income associated with financial autarky
after default. As with the value of foreign-currency borrowing, t D . N 13
The value of default when the economy enters financial autarky after default when
the sovereign has a LC contract is given by
def def
Vloc .yt / D maxŒu.ct / C.t /
N C ˇ.1 /Et Vloc .ytC1 / C ˇW .0; 0; yt /;
t
(18)
13. We have a counterfactual result that the economy without monetary commitment (original sin regime)
has a perfectly stable inflation rate (t D N ) in equilibrium. In Section 5, in which we extend the model
to include the Phillips curve, we find that the economy without monetary commitment suffers from higher
inflation than that with monetary commitment.Engel and Park Debauchery and Original Sin 1115
ct D h.yt /; (19)
where W(.) is the value of LC contract, which will be defined in the next subsection.
Downloaded from https://academic.oup.com/jeea/article/20/3/1095/6537405 by University of Wisconsin-Madison Libraries user on 17 June 2022
We also have t D .N
3.6. Original Problem under the Optimal Self-Enforcing Contract
We study an optimal self-enforcing contract in our model: The contract is optimal in
the sense that it maximizes utility of the representative household in the small open
economy. Moreover, the contract is self-enforcing in the sense that the government
under this contract does not have an incentive to breach the contract in any state of the
world. The optimal contracting approach allows us to model inflation targeting as a
committed monetary policy.
The original problem under the optimal self-enforcing contract is given by
1
X
˚ max 1 E0 ˇ t Œu.ct .s t // C.t .s t / /;
N (20)
for
ct ;btC1 ;btC1
loc
;t ;it tD0
tD0
subject to (1) the budget constraint, (2) the enforcement constraint, and (3) the expected
for
zero-profit condition for the lenders. .y1 ; b0 ; b0loc / are initially given.
it .s t1 /btloc .s t1 /
ct .s t / C btC1 .s t / C btC1 .s t / D yt .s t / C R bt .s t1 / C
for for
loc
; (21)
t .s t /
1
X
Et ˇ n Œu.c.s tCn // C..s tCn / /
N
nD0
n o
def
8st ; t 0; (22)
for
max V debase bt ; btloc ; yt1 ; yt ; Vloc yt ; ; for
X i .s t1 /
R D P r yt js t1 t : (23)
yt
t .s t /
Then, an equilibrium in this model is an infinite sequence of inflation and
interest rates on LC debt t (st ) and it (st1 ) in the contract, and allocations
fct .s t /; btC1 .s t /; btC1 .s t /g such that the contract and the allocations solve the
for loc
maximization problem subject to the budget constraint (equation (21)), the enforcement
constraint (equation (22)), and the lender’s expected return condition (equation (23)).
Note that the enforcement constraint equation (22) has two value functions on the
right-hand side: the values of debasement and default. The enforcement constraint
comes from two different types of limited commitment problems regarding the
government’s monetary and debt policy. These two enforcement frictions combine
to generate an endogenous debt frontier, determining the maximum amount of debt in
each currency. The debt frontier, in turn, affects the currency composition of sovereign
debt, which will be discussed in detail in Section 4.You can also read
